Name: Agentic AI Is Already Here - Learn how to automate the work you do every day with agentic AI
Uploaded: 2026-03-25T17:16:55.871Z
Duration: 44 min 8 s
Description: Agentic AI Is Already Here - Learn how to automate the work you do every day with agentic AI

Transcript for "Agentic AI Is Already Here - Learn how to automate the work you do every day with agentic AI": Alright. Good morning, everybody who's joined us. We've got a growing number of folks here. I'm Wendy White. This is Nick Lindauer, and we're gonna be your host today. We've published this as an office hours, and so we intend to keep it fairly interactive and answer questions. We've got a few slides we're gonna go through. Nick's gonna show a few demos of some cool ideas on ways you guys can get started with AgenTeq AI. And, you know, we'll we'll be open to questions all the way through. So drop your questions into the chat or the Q and A, and we'll try to keep your I will try to keep my eye on them and facilitate answering as many questions as possible as we go. If you are in the chat, we'd love to hear, you know, where you're from. So drop your name in and and, you know, where you're from. Let's let's let's see who's here. That'll help us tailor the content to you a little bit more. So, again, Wendy White. And why don't we just go to the next slide? Hi, Carmen. Good morning. So if you can drop your name and where you're coming from in the chat. I'm Wendy. This is Nick. And, you know, like I said, we intended this as an office hours. And, Nick, I think you just appeared off the screen. Come back. I didn't I didn't touch anything, so I think I'm still here. It's just popping in and out. I think probably set up to highlight whoever's speaking. That's fine. Good morning, Kyle, Laura, Frank, everybody. Thanks thanks for coming. We're gonna take you through a little bit about what Dexco's doing in AI for just a few minutes because I wanna make this as much as we can about you and learning. So I'm gonna go through that pretty fast, and then we'll get into the demos. And the idea here is give you some tools to help you get started, but you're welcome to ask questions about anything. So let's go ahead and move forward through the slides, and I'll just take you through them really quickly. So I don't know about you guys, but this has been the biggest transformational change in how I work. I've been in tech for thirty years, just to date myself. It's been the biggest operational change for how I work. I got my team started using AI about three years ago. But now, you know, it's to the point where it's a requirement for everybody in our team where every new job that we post, we're interviewing for AI skills. And the folks that are here, we're trying to upskill as fast as possible. So learning how to use both generative and agentic AI is, I think, going to be a incredibly common skill for everybody. So no surprise to those of you who are here who are probably worried about it. You know, it's moving so fast. Sometimes it's really hard to keep up. So if you're feeling stressed about it, you know, remember that, you know, only a small percentage of of companies and people are still are are adopting it so far. And so you're probably not as behind as you think you are, but this is gonna be a good opportunity for you to catch up today. So next slide. Nick? So as we're we've the last, what, two or three years, it's been generative AI. It's using chats, using Cloud to create content, do text, videos, assets. That's generative AI where it actually generates the content you're asking it for. You're giving it some prompts. You're giving it some context, and it goes and does that one particular task. Agentic AI is where we're moving into where AI agents are now gonna run your workflows and take actions based on the results of those actions and those those parameters you've given it. So you're creating robots that will go and do all these actions for you without you having to do multichain prompting or what you had to do with generative AI. And generative AI still has a strong place in the market. But now moving into Ingentic, we can a lot move into the autonomous workflows that these agents can take over. Like, some of these day to day tasks that we don't need to be doing or spending our time on, you just need the results. This is where EngenTic AI is helping us level up our game. Next slide. So I actually just kinda talked about this. Right? With AgenTek AI, what we can do is we you know, we automate more marketing. We can go through and pull in member information and create emails based off of a certain aspects and things that they've done. We can collect more member behavior, what they're doing on your website, what they're doing within your member management software. If we're uploading local signals intense signals, we can work on, running platforms and, agents off of those signals, and then we could scale. Gyms can use a AgenTeq AI to scale both custom coaching, content, operations, whenever you may need to scale, that agent can come in and actually help you and your team go from, you know, a four person team to a a 10 person team without adding any additional headcount, which is huge. And we can we can reach more people quickly with this with agents and generative AI combined together. Yep. Next slide, please. Okay. So what does DAC's go up to today? You know, we are thinking really hard about where to responsibly use AI in a way that is keeping your data secure and compliant. So I'm just gonna show you a few things about our approach, but we're literally building it into every part of our member management platforms and our marketing automation platforms that clubs count on to operate today, identifying all the places where we can automate workflows to reduce, you know, staff overhead and complexity, but as well as to help you scale and grow your business. So next slide, please. So just a few examples of some of the AI native capabilities we're building into into Doxco. I'll talk about the MCCP server in a minute, but we're we're doing things from leveraging the data that's in your system to help you do things like predictive intelligence around churn prediction or lead lead scoring or even capacity planning in your clubs. In the middle there, you can see we're doing some workflow automation, like scheduling optimization, you know, billing resolution, membership processing, digital ad testing. Over on the right, we're gonna be launching our own AI agents for conversational AI, for that AI front desk, AI support, more lead follow-up automation and Engage Pro. And then, obviously, like, you know, all of you are going to need to interact with your customers, and we need that conversational voice so they can call in, they can call out, help you grow your business through those kind of agents. All of these things are on our road map or are being delivered in our products today. And so if you're interested in that, you can obviously ask for more information. But back to the point of today, which was office hours help you get prepared. Wanna make sure you know what you can do. So next slide. So, Nick? So looking at this, when we're talking about we we've we've heard about generative AI. We've heard about agentic AI, and now you're talking about MCP, which is the abbreviation for model context protocol. That is a system that connects all of your AI systems to your systems. It's an orchestration layer between the LLM and your systems. It allows the LLMs and the the AI channels that you're using to access your data and understands the context. So the MCP sets up that context, and it's the structure to allow everything to work together and actually speak the same languages. So we can layer in customer data, unified it into a unified data layer, which then becomes goes into the MCP, where it's then goes into Daxco native AI, any approved external integrations. We'll talk a little bit about some of those integrations and what you wanna do there, and then the partner AI with Daxco exchange. So we make sure that that you have everything governed, everything is audited that connects into that MCP, and everything is permissions. So we don't we don't have any of these horror stories that we've heard about data leakage or, you know, chatbots giving away information, anything like that. Yeah. So, one, we wanna introduce you to this concept to tell you it's coming from us so that you can have some choice about what AIs you use. Maybe your company has already, you know, picked a voice AI agent, and you wanna connect it deeper into your Daxco systems. This is how you're gonna be able to do that and get, you know, secure access to your data to run AIs that may not be offered from DaxCo. So we wanna let you know we're we're working on this so that you can, you know, obviously, all the DaxCo capabilities, but that we also wanna have, you know, choice and flexibility in our ecosystem and let you bring your own own AIs. So let's move on because we want we really wanna do is get you to the point where you're feeling, like, confident building and using some of your own AIs. Next slide. Do I show the marketplace, Wendy, or you wanna talk about it? Yeah. I'll talk about it for for a second. So for those of you who are Club Automation customers, if you go to clubautomation.ai, you're actually gonna see the marketplace that Nick is gonna show you right now. For those of you who are Daxco customers or Daxco operations, that AI marketplace will be available over the next week or so. So the reason why we put this together is we wanted to give you some tools, some templates, wanted you to see every single AI that is on our road map right now or the capabilities coming, and then also have one stop shop for you to look at what AIs, the partners through the Dexco Exchange ecosystem are doing. So Nick's gonna show this to you, and then he's gonna click you through, like, how to use some of these agents and how to set them up and play around with them so that you can get started doing things today. So, Nick, I'll let you Avi, stop sharing these slides, and I'll let Nick share his screen and take you through it. Good. Let me get over the entire screen. Alright. So as Wendy mentioned, we have let's make sure that shares. Good. We have clubautomation.ai. This is hot off the press as of, I don't know, probably about ten minutes ago. Was working on this, making changes, and updating. So you see that we've got this layered out where we have agents. We have agent capabilities. There are layers that we can hear that you can break out. So, for example, if we look at our member newsletter writer, this can be an agent, but we've given you all the prompts to just start using this, like, to see what the output could possibly be. So you can come in here. You can download all these prompts. You can copy and paste these three here. If you actually sign up and try the free sample, I think you get, like, 12 or 15. They're all custom for clubs to use. Those are just prompts. That's not the actual agent. Right? The actual agent has to be built and customized to your workflows, your systems, depending on if you're using Cloud, chat, or however you wanna run this through, you know, either make or n eight n. So that's the layer that we have to help you with, but I wanted? to make sure that we gave you prompts. Yeah. Can we pause for a second? Trudy's saying she's not seeing the screen. So is anybody else having that issue of not seeing the screen? I see it where I'm looking. Okay. Garen says he's he's seeing it. So, Trudy, maybe you need to pop out a different screen. I'm not quite sure. But if anybody else is having oh, thanks, Frank. Never mind, Nick. Keep going. Okay. And, Trudy, if you're not seeing it, message me, and we'll try to get some support for you of maybe what's going on. Alright. Keep going, Nick. Yep. Sorry about that. Yep. So this is very, very simple. You can come in through here. You can try the free samples from each of these agents. Each of these agents has a free sample prompt, actually three different prompts that you can try to just see what that output could possibly be. Because we wanna make sure that, hey. If you can't touch it and feel it because it's not customized yet, you can at least start to see what that output could be when it is customized to you using your data, your information, and on your platforms. Right? And as you come through here, you get all sorts of sorting, free pro coming soon. So a lot of our coming soon stuff is what Wendy was talking about. We have a unified comm center coming soon with Engage Pro, making sure that you can see all of your customer communications in one place. That is coming soon. We also have layered in here, all of these prompt behaviors that we can work with with Engage Pro. We have our partner ecosystems in here. So if you're looking for a partner, you wanna layer in the AI that they work with for, buy or keep me, you know, depending on what you're looking for for them, you can get into those and layer in that those pieces of AI. So this is the hub that we're creating so that you can see agents that are independent of our software. You can see the AI automations that are coming into the software or are already in our softwares as well as our partner's software that interacts in the AI capabilities. Anything else we wanna show here, Wendy? No. I think get into, like, giving them some giving them some, you know, how to's. Okay. So I guess from the the crowd, we didn't do a poll, but can we still see my screen? Good. So what I wanted. to start with was the word of the day, Claude. We we both we have enterprise chat GDP, and we also have enterprise Claude here. I wanted to start today with just general prompting so that we can see the differences in prompting, and then we'll go we're gonna go from prompting to some deeper research, and then I and then I did build a spin up a real quick agent so we can take a look at this. So we all know prompts. You you go into any chat or cloud, and you just say, hey. Write a promotion email for my gym. And I've got two here because I wanted to show you the differences between the two. On the left, it's just a bare bones prompt. On the right, I've given it a little bit more context, but still not great. On the left, it's struggling. It's asking for some more information in some cases, but you can see the differences between these two outputs. I didn't tell it what gym I was. I didn't tell it anything about me. I didn't tell the offer, so it just started making stuff up. That's where you get AI slop is. When you don't have the right context or the information to give your generative AI or your agents, you're gonna get a lot of slop. Now over on the right, I gave it a little bit more information, and it it spit out some stuff. You know, it's it's trying to tailor it to Little Rock. It did I did not give it any detailed information on Little Rock, but I just gave it some. And it's it's pretty long. It's not, you know, where I would necessarily want to have that for an email. So if I pop over and customize that just slightly more let me get to my prompts over here that I saved. And I'm gonna take this one. And now I'm gonna say, hey. I'm now on the marketing director. Now this is we should all be pretty familiar with this level of prompting. We're giving it context. We're telling it who it is, what it's doing, and what we want it to do. But I've also given in the piece of the role, the context, the task, and the format. Now we can see the big difference between these two emails. So this one's very generic, and this one's custom and tailored to that exact use case and what we want. This is generative AI. This is just prompting, and this is just how to do a better prompt. You can do the exact same thing in ChatGDP. I'm not saying do Claude or ChatGDP. I just pulled this up for an example. Any questions about the prompting? Not seeing any questions or any q and a, so. I think we keep going. K. So the next level to talk about is going into projects. So projects are where you give Claude or ChatGDP instructions, and you and it contains the output within those instructions, and it helps you tailor the output with those Again, we're still in generative AI, but we're gonna be looking more about, okay, how we can get a little bit better. So in this window where I'm still very general, I'm gonna give it just the same generic prompt. And if you've noticed here, I'm telling it to ignore any writing skills because I've uploaded some custom skills with my writing language and how I write. So I'm telling it to ignore just so we can get the basic output. But so that's a generic one. And then I did spin up a project for Little Rock, and I'll show you those instructions here in a second. But I'm just gonna give it that exact same prompt, and we're gonna compare with instructions and without instructions. Oh, see, it's it's trying to pull in my master writing skill. So even though I told it not to, it's trying to do that. So you have to be mindful that it's it might actually get into your writing skills. But we'll see here it's doing the same thing over here. But now it's using the Little Rock voice guidance that I gave it. So it's pulling that together. This is getting better over here in the generic space. But now it's hey. Here's this is pulling in Frank and Mary and the l r LRAC team. It's asking them to stop by the front desk. It's a much more engaging email, and I didn't do much except I put this within a project. And then the instructions that I gave the project, I gave it the role, I gave it context, and I gave it some information and tone. And so those were things that were missing from that first couple of prompts. And you can honestly, I use Claude to make my instructions. So I'll go into Claude and say, hey. I'm creating a project about this. Help me write the instructions. And then I use I paste those in there. So it's it just kinda builds on itself. So the idea is that a project, you can keep keep going back to it, and it has the training information in the project. So. if you're working on something and you're doing it all the time, if you rather than reprompting it, re reminding it, giving it instructions every time, you just keep that in the project. Exactly. And then what you'll also see here, I just asked it for another one. It's gonna run that in the background. But you'll see here, it creates a memory. So that's what Wendy's talking about. It creates a memory that of what you're doing within that project so that it constantly understands and refreshes with every single new chat you put in there. It has instructions, and then you can upload files. That's the exact same system that you have in ChatGDP. So it's not a different system. They've actually become basically on parity with each other. So everything we're talking about for Claude, you can actually see in ChatGDP. One of the things, though, you can save if if in a project, it starts to create a file for you, like it's created a recommendations doc or it's done something for you, you can save that file to your project. That way, you can always go back and reference it. So that's one difference in Claude. You can't do a ChatTP, but you can always upload it into that project in chat. Okay. So let's just recap. The first set of prompts that Nick showed you are just prompts, and now Nick is showing you how to create a project and keep that project as a fixed project that goes forward inside your cloud. It works the same in ChatHVT, by the way. You can create a project there. And same thing, prompt that project, store data in there, and then it will remember all of that. And it just gives your your generative AI we're still in generative AI. It gives it some, you know, better context for you to get better results. Yeah. And so the next one is going into AgenTix. So I I built this artifact. So this is a it's an agent, but it's also a Claude artifact. I built this in Claude. All and I did it this morning. I I went through the list of people who are attending, so I'm hoping hoping Little Rock and Monroe are here. But what I did is I said, hey. I wanna build a competitive intelligence agent. This agent, you can see I put in a graph so you can just visually see what this agent does. The CEO or somebody comes in here and selects a competitor, then the agent gets the goal. It goes through. It does the searching on its own. We're not telling it to do anything. We're just selecting who it's gonna search for. Then the agent will actually come through and synthesize the findings for us and create that intelligent report. What I would like to see on this, like, if I was gonna take this to the next level, I would send this report into a database. I would send it into Slack. I would notify the team. But just for the purposes of this demo, I just did a quick quick and dirty agent. So in this case, we're gonna say we're gonna go with Crunch. We're gonna hit run agent. I put in here a visual so we could see it, but I don't wanna make sure that we have I don't wanna have any dead time, so I also pulled up the results already. So this is the results when we ask it to look at Crunch. I could put any competitor name in here, by the way, and it's telling me exactly what's going on with Crunch and the Monroe family Monroe family YMCA. What's happening. in the market? such an expert here. You flew through that, and I felt lost. So I think people are on the phone are gonna feel. lost. So let's go back to the beginning and talk about like you said, this is an artifact, but it's also an agent. Maybe just stop and go back and say, what did you do here? Okay. So what I actually did I'll pull this screen over here. Let's see. I am in Claude. So I use Claude desktop. I you can see I'm also using Claude dot a I. So if your brain's starting to hurt, just just know that I'm running four different computers at the same time some days. But I'm using Claude desktop. And what I just literally asked is, hey. Help me create a competitive intelligence agent. I want I'm representing Monroe. The and this is this is multiple iterations that we've gone through, in the last couple of hours. And then it spins up an artifact. So Claude will spin up an artifact that then you can then share within your company, or you can share it publicly. So this is the artifact. It's and then you can see here, I can share that artifact. That's what I shared out. That's the URL that I'm using. So I can hit that from anywhere in the world, and I can start to use it. It this doesn't contain any proprietary data, so it's it's free to be opened up, and that's that's okay. to build it here. But this is a. this is a basic agent. Again, it's running through things that I'm not asking it to do. It's just doing that versus me having to prompt every single thing that it does. Yeah. So in this instance, you're the CEO of of the YMCA in Monroe, and you've asked it to assess who's nearby that are that could be competitors to your why. Correct. And now you're going to run a an an an agent that tells you how to compete with them. Yep. And it's gonna it's. this particular agent is looking at pricing hours. It's pulling in recent reviews. It's looking at their website. It's looking at anything that's out there in the social media space that the competitors might be doing. And as this runs, you and you can't really see it right here, but I can I can put it in any competitor here? What it did when I did the setup, it said, hey. These are the most popular ones. Come to find out anytime fitness is closed, but it's still in the search results, which is interesting because from a competitive standpoint, they're not stealing your business, but they are stealing your search results and web space. But then the results. Right? This is it's not perfect. I was working on it this morning, but it broke it out by marketing and positioning, telling me, like, Crunch reviews are very, very clear that it's a honeymoon period. It's a new facility, but the there's a ton of complaints about billing, cancellation, and overcrowded facilities. Like, that's that's customer sentiment is not great. But it it indicates that the y has a very strong community loyalty in Monroe, and Crunch has is facing those challenges. So this, is. You're getting a lot of free work today. yeah. Nick was gonna share this with you after, which is fantastic. And then I I actually did ask it to break it out into programs amenities to make sure that we're looking at, hey. What are the comparisons between each of those competitors? And it's giving us some really good, at least, starter level marketing insights on how you should run a campaign versus Crunch if you're thinking about that. Again, I'm coming into this cold. I'm not the expert in Monroe. You would definitely need an expert within your business to to work with this and say this is where we want it to go. But it gives us a lot of detail, especially around, you know, thoughts and where we should go with this. And we can take this and customize it further. Okay. So the difference between the generative AI and the agentic AI here is go back to the artifact and the prompt that that Nick told it to, it actually went out and got secured the pricing. It secured the reviews. It pulled in, you know, all of that information on your behalf and then analyzed it. You told it, go pull that information, from there, build a competitive analysis. How. should I compete better with that with. that, Jim? So, again, it's doing all this work for you in the background. You're not having to ask, what are the reviews? Right? Or tell me the insights about the reviews. You're you're telling it, go do this assessment. Go build this thing for me. And so it's doing all those steps for you in the background in a few minutes that. would have, you know, previously taken you hours to go poke around, probably go walk over to that facility, go look at their website, go, you know, spend time on their review page, and then it's doing some thinking for you and coming up with some ideas for some campaigns. Right. And then the other piece of this is is you spin this up as an agent or an artifact or however you spin it up and you distribute it to your team. Every time they use it, the results, they're gonna be they're gonna match versus you're telling them that, hey. Use a particular prompt, and then they may go off and change that prompt slightly, and they get slightly different results. They say you've given everything. You've given your agent the rails, the systems to connect into, and so the results are persistent, and they're repeatable. And then you can you can layer on this. So as you start to build a database about competitive intel, you can then pull competitive intel and actually run battle cards and, you know, customer insights and all kinds of things off of that so you can layer on other agents about off of the output of this particular agent. Yep. I think, Nick, what wasn't clear again is how do I build the agent? What's the difference between writing a prompt and building an agent? So maybe just walk through those steps of, like, what's the prompt or the action you have to tell Clog for it to make it an agent versus just a generative AI prompt? Very simply, you ask it to say, hey. I'm able to wanna build an artifact or I wanna build an agent, and you start telling it what you want to do for this base level of an agent. Right? You're just you're you're working with it. Hey. I want an interface. I want it to do some automations on its own. You just ask it to help you create an agent, and it'll it'll start. to spin that up pretty quickly. Now we do have very complex agents and workflows that go beyond that. Right? So this. is for purposes of demo. But I think I think the key thing for all of us is don't be afraid because it if you've been using GenerativeEye, going one step further to an agent is just talking to your your Claude or your Chad, as I call Chad GBT, talking to talking to them and asking them to help you build an agent. It's it's it's basically your brain combined with the AI. Yeah. It it isn't some, like, special skill you need to have. You need to tell it, we're gonna build an agent, and it will help you build an agent. Now what version of Claude or ChatGVT does somebody need to have for it to be able to do the agentic workflows? What what what level? So in this case, I I mean, I I have the most recently updated version, but I was actually just using chat for all of these. I wasn't even using code. So I just came in here and said, hey. I need to do a demo or help me run a chat, and then I created a project. But this is all in chat. So you can go into Claude code, a deeper. it's a basic, yeah, it's a basic paid version of Claude. Yes. You don't need. anything more than the basic paid version of Claude or or ChatVT to do this. Correct. Okay. Yes. Alright. I'm gonna turn back to the audience and ask for questions on that example, more of, like, how do you talk to Claude or ChatBeti? Does anybody have questions? Or should we just move on and have Nick show us another example so we can really wrap our head around it? I don't see any questions popping up. I haven't seen a comment for a few minutes, so I'm hoping you're all still there. And if you're alive, give me a high five in the chat. Marcy says, more examples, please. Let's move on. Let's go get that next. Well, let's let me share my where is the button again? Wow. It's hard to find. I'm gonna share my screen. We didn't talk about lot of pages open. we didn't talk about skills, but skills are. a layer. Let's talk about that. They are also available in ChatGDP. So be let me make sure my screen starts to show. There we go. Skills are something that when you create a skill in Cloud or ChatGDP, it is then accessible by every chat, every project, or every Cloud code, or every instance that you use can actually pull that skill. So skills, I I like to think of them as an umbrella. And everything that you do can access that skill. And you saw, like, sometimes I have to tell it, don't access that skill because I don't want you to use that skill. Sometimes you may have to call it and prompt it and say, hey. Please use this skill because it may not know. Fortunately, if you're doing, a writer skill, like, that's it's pretty common that it'll it'll pull that. But you can see here, like, my skills, I've got copywriter skills. I've got context architecture. I've got some safe harbor agents. I've got b to b copywriter. So it does get a little confused sometimes, so I do have to specify which skill to use. Skills are very easy to create. It's the same thing as, like, what we were talking about to create an agent. You just go into Claude, and Claude has this skill creator as a default. So you can go into let's see. I'll just walk you all through. You go into settings. You click capabilities. And for some reason, they put it all the way down here in the bottom. You go to customize, and it's your skills. And they have example skills already built in that you can just go and turn on. I recommend turning on skill creator because that's the tool that will help you create skills. That's a skill for skills. And you just then go into your chat, Cloud Chat or ChatGDP, and say, hey. Help me create a skill related to this. And then Take a step back. again. So we talked about prompts, then we talked about projects. Right? Then we showed creating an agent, and now we're talking about skills. And I think the the thing here about a skill is it's a perpetual skill that you can use anytime inside any project or outside any project. And it's a a a set of directions and activities that that help build the skills. Yep. Are you using AI to help build the skills? You bet, Yes. Frank. 100%. So why don't you why don't you show us how to build a skill? Yeah. Actually, I just built this one this morning. So we needed we're having some PowerPoint templates that needed to be make sure that they were, you know, reined in on our brand design. So what I did with that, I went to my favorite buddy chat, and I came over here and said where was it? Maybe I lied. Maybe I created it as a project because I'm pretty sure I was like, I'm gonna have to do this again. So I did. I lied. I had a project created for it, but I just used chat. I went in and said, hey. Let's make sure I need to make a skill that actually designs PowerPoint. Images and brand materials using our brand standards. It already knew our brand standards, and then I uploaded the patterns, the files, and the imagery, and I created a skill, from that. And so it started working through that, came down here. It created the skill MD file. It included all the references. I went through it, double checked it, looked at it, and then it's you can download it, share that with your team, copy to my skills. I asked it to create a ZIP file that I could share out with everybody. And here's the full architecture, and it downloads it, and it's ready to go. So I can start to come in here and create branded PowerPoint templates or branded one sheeters that match our DAXCO brand within Claw without having to upload anything. Yeah. So now you can share that skill with everybody else in your team that has Claude, and they can save that skill. And next time they want to create a presentation, they can go to Claude and say, leverage the Daxco PowerPoint skill and build me a presentation that says a, b, c, and it kicks out PowerPoint in the brand. Correct. And it will also, That's useful. references my other skills. boss wants to know Your boss wants to know why you didn't give her that skill already because I could use that today. it, like, five minutes ago. Alright. Like, could. you it doesn't have a time stamp on it, but it was literally, like, a couple minutes ago. But we're testing it. But you can see also. what the skill does is it references my other skills. So it's also saying, hey. I know that I have a brand design system. I know I have a PowerPoint skill because, Claude, you can enable certain capabilities. You can enable PowerPoint. You can enable Excel. And it's also layering in. It's gonna reference my master writer because I've given my Claude, hey. Every time you write, you're writing as Nick. This is Nick's master writer skill, and then I have a copywriter skill in there too. Okay. Alright. Frank and James had some questions about when are these capabilities coming to CA or Daxco or any product from Daxco you might be working with. So there are some ChatchPT interfaces today in our products, and more and more will be coming soon at MCP server. We're doing a road map webinar quarterly now where we're going through the new releases, the new capabilities in the products. So I think our new our next road map webinar will be mid April, and we'll show you the timeline and exactly where you can when you can expect certain capabilities and skills like the MCP servers or additional prompts inside the product, including, like, when when can you prompt your data for reporting, etcetera. So all that will be coming in the next road map webinar. But, also, you go if you're on CA, if you go to that ca.cloudautomation.ai page, you'll see you'll see the list in there, and we'll start updating that with exact dates as the dates come to us from engineering of when we can expect those. So two ways for you to know is always check back on that clubautomation.ai page. That will be consistently updated with dates as we get them. And then two is watch out for those webinar invites so that you can get, you know, a detailed overview and see what's coming in the next ninety days from our product and product marketing team. Alright. Any other skills you wanna show off today, Nick, or or or should we wrap up and see if there's any other questions? I'm happy to keep showing off, but if there's custom questions, people are like, hey. What about my instance? Happy to do that too. Someone says, thank you. Looking forward to the growth and opportunities. Me too. AI is exciting. It's also, you know, a little bit overloading. And but the more we use it, the more that we, I think, can all be better about our business. Okay. Rebecca says show one more. Let's go. So what else do you wanna say? I I mean, I'm learning a lot. I don't know about you guys, but I'm learning a lot. I I really love projects, honestly. So I spend a lot of time in projects. One of the things actually, you know, we had an HR conversation the other day where they said, hey. Do we need this database? And I said, actually, no. You don't. What you need because they what their challenge was is they were trying to make sure that their job descriptions were constantly referenceable. And there is a limit. I'm not sure if folks know this, but there's a limit in ChatGDP and in Cloud with file uploads. So if you're trying to do a ChatGDP project, I think it's 20 files or some sort of space capacity. The workaround for that is to create MD files and absolutely use Claude or Chad to do this, and that you you're just literally telling it to create an MD file. An MD what is a it's a markdown file. So it takes out all of the marketing fluff that we love, takes out all of the extra words. It takes out everything, all the images and stuff, and it is literally a text file that tells the agents only exactly what they need to know. I'm gonna try to pull up one here, but that makes it easier for the agents to parse your data. So if you want it to run through 20 job descriptions instead of trying it to run it through 20 PDFs, upload those PDFs once to chat and say, hey. Create an MD file of these job descriptions, and it'll come out, and it'll spit out one file that you then upload into your project. And then it can reference that much quicker and actually with better clarity and data reporting back to you. I think one thing that I have learned and as may maybe some of you are going to start using AI and buy, like, an enterprise account, you'll start learning, like, about tokens. One thing that Nick was talking about is if you strip out everything and have an MD file, it it's a lighter load. It reads faster. But if you are actually starting to do more complex stuff and you have an account and you're using tokens, it actually uses fewer tokens well. So that's a cost saving thing for you. So you might wanna think about that if you're gonna be doing it in your own AI projects. Alright. Someone's asking, how is the AI being implemented in CA analytics? Is it already compiling data, and and how is that data? So I'm sorry to say we're gonna have to bring that one to the product team. I have a general and good idea on that one, but not an exact one. So I think that's a really good question for the product road map webinar that we're we'll be having in a few weeks because they'll be able to explain exactly how the AI will be exposed on top of the data the the data vault in Snowflake and how you'll be able to query it. So sorry. Can't answer that one. I am I I'm almost guaranteed to get it wrong. I have a general idea, but not enough to wanna be the expert on that. So if you message me outside, I'll bring that in. Is club automation available now? Yes. You should be able to find it. You go to clubautomation.a find what Nick was showing. I would have a bunch of the example prompts that he talked about. Yeah. And you can download those prompts. They'll be emailed out to you. Or you just go through and copy and paste them and just start working with prompts. Yeah. Yep. Alright. Nick, one more example, or are we are we are we are we done for today? Are we I think we're I can Just think we started talking about the MD file? So let me show that real quick just to tell people. It's boring. It's nerdy. It's just it's a it's a plain file. But in this case, this is an MD file for an agent that we're working on. Wendy knows this, but it's not ready yet, so don't don't share it out, Wendy. But this is the it's a it's a markdown file for an experimentation in CRO agent. It's very boring. Right? But if you read this, it has exactly what the agent needs and nothing more. And that's the that's the clarity that you get with creating MD files is just very stripped down instructions for your agents or your projects to work with. So a lot of times, what I do when I'm working on a project and I'm going multiple files deep and all of a sudden, like, I don't know where I am and I don't know what day it is. I go back and I say, hey. To Claude, please create a MD file of where we are with this project. That way, I can retain that knowledge. So you can create these markdown files for instructions, but also memory and references so that you can come back to where you were with what you built on. Because sometimes you'll get really, really deep into something, and you've gone down some rabbit holes, and you need to actually go back to a a staging point. Okay. Very good. Alright. I'm gonna say thank you. If you liked this office hours and you want us to do it again, we're happy to do it, you know, once a month or so and share either new skills that we've implemented, new agents we put on the dot the a I sites. And, again, for those who are on the DAX Corporation side, we'll get that one launched really quickly. But we're happy to do this. We're learning every day. These are the kind of, you know, experiences that Nick is training our team on. Nick is our official VP of of marketing AI for my organization and builds all the agents and teaches everybody. And so a lot of the agents you're gonna see on a CA club automation yes. A sandbox session, that would be fun. We should definitely do that. Happy to do that and happy to bring our product team on next go round so they can answer product questions as we go too because I think that Yeah. a lot of things are gonna spark ideas about how you wanna work with the data and the product, and we'll and we'll do that. So look for another office hours probably in about three to four weeks. Show up in your inbox. We'd love to have you come back, and we'll do more of this. Look for that road map briefing coming to you in April. And if you have any questions, don't hesitate to ask your account team, and they will definitely get it in front of us. And if you need any other thoughts or advice, Nick is just [email protected]. I'm Wendy White at daxco dot com. We're happy to help out as well. Alright. Thanks everybody for joining today. Appreciate you. Bye, everybody. Bye.